In processing a mesh point cloud, the processing speed thereof is in a inverse ratio to a number of points and grids of the mesh point cloud. A mesh point cloud may contain too many points and grids that it requires a relatively long time to process it, for example, to model or transmit the mesh point cloud. To reduce the number of points and grids of the mesh point cloud in order to reduce the required time for processing it while not scarifying the geometrical precision, likeness and characteristic of the mesh point cloud becomes an issue. Presently many systems and methods are proposed to solve this problem. However they only consider the geometrical configuration and topological information of the mesh point cloud, without considering the problem of closeness (completeness) of the mesh point cloud after it is simplified. Thus, the effectiveness of the simplification is not satisfied since the geometry of the simplified mesh point cloud is deviated too much from the geometry of the original mesh point cloud.